FEBXUARY 1967
XOTES
1, X=CH==CH 2, X = CH,CH,
\
/
CH, 3,X=CH=CH
473
heptene (5) was converted into the chloride by heating the mixture for 5 min with thionyl chloride and removing volatile material under reduced pressure. A suspension of 4 g (0.02 mole) of sodium p-toluenesulfinate dihydrate in 50 ml of ethanol was added to the chloride and the mixture was heated for 30 min a t reflux before being poured into ice-water. The solid was collected, washed with water, and dried, giving 1.5 g (90%) of sulfone 3, mp 210-211". B y using the same procedure, p-tolyl 1OJ1l-dihydro-5H-dibenzo[a,& cyclohepten-5-yl sulfone (4) was prepared from 5-hydroxylO,ll-dihydro-5H-dibenzo[a,d] cycloheptene ( 6 ) ; the yield was 0.4 g (23%), mp 187-188'. Reaction of 5-Hydroxy-5H-dibenzo[a,d]cycloheptene with pToluenesulfinic Acid.-To a solution of 1.0 g of sodium ptoluenesulfinate dihydrate in 25 ml of ethanol and 2 ml of acetic acid was added 0.5 g (0.0024 mole) of 5-hydroxy-5H-dibenzo[a,d]cyclohept'ene (5). After the solution had been heated a t reflux for 30 min, the solid was collected; the yield was 0.70 g (85%), mp 210-211'. A mixture melting point with sulfone 3 was not depressed. In a similar manner, 5-hydroxy-l0,11-dihydro-5II-dibenxo[a,cl]cyclohept~ene ( 6 ) gave a 15% yield of p-tolyl l0,ll-dihpdro5H-dibenzo[a,d] cyclohepten-5-yl sulfone (4), mp 187-188". A mixture melting point with the authentic sample was not depressed. Reduction of 5H-dibenzo [a,d]cyclohepten-5-one with Anthracene-g,lO-biimine.-A mixture of 1.0 g (0.050 mole) of ketone 1, 3.0 g (0.15 mole) of anthracene-9,10-biimine,6and 50 ml of ethanol was heated a t reflux for 1 hr. After cooling to room temperature, the anthracene was removed by filtration and the filtrate was concentxated and chromatographed on alumina (20 9). Anthracene and starting ketone were removed with benzene, and alcohol 5 was eluted wit,h ethanol; the yield was 0.25 g (237,), mp 118-120". A mixture melting point was not depressed, and the infrared spectrum was identical with t,ha.t of an authentic, sample. 5H-Dibenzo [a,d]cyclohepten-5-one p-Toluenesulfonylhydrazone.-A 3.7-g (0.02 mole) portion of tosylhydrazinc was added to a solution of 5,:i-dichloro-5II-dibenzo [a,d]cycloheptene,* obtained from 4.0 g (0.020 mole) of ketone 1, in GO ml of acetonitrile. The mixture was stirred overnight and the solid was collected. Additional material was obtained by concentration of the filtrate, and the solid was recrystallized from ethanol-acetonitrile to give 5.2 g (70%) of the tosylhydrazone, mp 213-215" dec (lit,.4 mp 204"). The nmr spectrum is in agreement with the proposed structure. Anal. Calcd for C?~HIRNZO?S: C, 70.6; 11, 4.8; S , 7.5; S, 8.6. Found: C, 70.9; H, 4.6; N, 7.7; S, 8.6. The tosylhydrazone (1.0 g) was recovered quantitatively after treatment for 16 hr a t reflux in 20 ml of ethanol. containing 0.5 ml of acetic acid.
[f?qJ 4, X = CH,CH2
H
OH
5, X = CH=CH 6 , X = CH2CH,
thermal decomposition of t~sylhydrazine.~This reaction has also been shown to produce p-toluenesulfinic acid, the other required reactant. Reaction of alcohol 5 , obtained by reduction of ketone 1, with ptoluenesulfinic acid was shown to produce sulfone 3 under the original reaction conditions. Alcohol 6, obtained by reduction of ketone 2, behaved similarly and gave sulfone 4. I n support of this mechanism, it was found that ketone 1 is reduced to alcohol 5 when heated in alcohol with anthracene-9,l0-biimineJ6 a source of diimide. The reduction of ketones in this manner may occur when the carbonyl group is sufficiently inert to nucleophilic attack by tosylhydrazine to permit diimide formation by the thermal process. An alternative mechanism considered was formation oi the tosylhydrazone, decomposition to the diazo compound, conversion to the carbonium ion in the protic medium, and sulfone formation with p-toluenesulfinic acid. This is ruled out because the tosylhydrazone, prepared from 5,5-clichloro-5H-dibenzo [a,d lcycloheptene, is recovered unchanged when subjected to these reaction conditions. Experimental Section7 Reaction of 5H-Dibenzo[a,d]cyclohepten-5-onewith Tosylhydrazine.-A mixture of 5.0 g (0.024 mole) of ketone 1, 7.0 g (0.038 mole) of tosylhydrazine, 100 ml of ethanol, and 1 ml of acetic acid was heated at reflux for 15 hr. The solid that separated was collected and recrystallized from ethanol-chloroform to give 3.4 g (39%) of sulfone 3, mp 210-211'. Anal. Calcd for CZZHI~OZS: C, 76.4; H, 5.2; S, 9.3. Found: C, 76.5; H , 5.3; S, 9.3. The nmr spectrum (DMSO) had peaks a t r 2.6-2.8 m (aromatic), 3.37 s (olefinic), 4.12 s (benzylic), and 7.93 s (methyl) in an area ratio of 12:2:1:3. The infrared spectrum (KBr) showed strong sulfone bands at 1310 and 1140 cm-'. A mixture melting point with an authentic sample, prepared as described below, was riot depressed. When 5 g (0.024 mole) of lO,ll-dihydro-5H-dibenzo[a,d] cyclohepten-5-one (2) was used in place of the unsaturated ketone 1, 1.1 g (137,) of the corresponding sulfone ( 4 ) was obtained, mp 187-188'. A mixture melting point with an authentic sample, prepared as described below, was not depressed. Anal. Calcd for CZ?'HZOOZS: C, 75.9; H, 5.8; S, 9.2. Found: C , 75.9; H, 5.9; S, 9.0. p-Tolyl 5H-Dibenzo [a,d]cyclohepten-5-yl Sulfone (3).-A 1.O-g (0.0048 mole) portion of 5-hydroxy-5H-dibenzo[a,d] cyclo( 5 ) R. S. Devey and E. I:. vanTamelen, J . Am. Chem. Soc., 83, 3729 (1961). (6) E G. Carey and W . L. Mock, ibid., 84, 685 (1962). ( i ) The nmr spectrum mas measured on a Varian Associates Model A-60 spectrometer and the infrared spectrum was measured on a Perkin-Elmer Model 137B spectrometer. h4elting points are uncorrected.
(8) J. J. Looker, .I.
Om. Chem., in press.
Bicyclic Enamines. 111. Reduction of Enamines with Secondary Amines1V2 A. GILBERT COOKAND CHERYL R. SCHULZ Department of Chemistry, Valparaiso University, Valparaiso, Indiana
Receitied August 2.$, 1966
The reduction of enamines derived from both bicyclic and monocyclic ketones by secondary amines has recently been rep0rted.l The only secondary amine used in these reductions, however, was hexamethylenimine. The reduction of enamine intermediates by methanolic solutions of either dimethylamine or piperi(1) For part 11, see A. G. Cook, IT.C. hIeyer, IN=C< stretch) and its gas-liquid partition chromatography retention time with an authentic sample.' The corresponding imine oxidation product of unsubstituted hexamethylenimine itself resulting from the reduction of enamine 3 could not be isolated because of its known instability.' This undoubtedly accounts for the large amount of dark, intractable residue found upon distillation of the products resulting from this reaction. N-Benzyl-Ncyclohexylideneanilinium perchlorate ( 10) is reduced by N-benzylaniline to yield Y-benzyl-N-cyclohexylaniline (11) and benzalaniline (12). The latter com-
0
c10,5
4
6
5 in an 85% yield.' It would be expected that the stereochemistry of the product obtained from lithium aluminum hydride reduction of a bicyclic iminium salt such as 4 would be endo as has been observed in similar reactions where the exo side is the least hindered side for hydride a t t a ~ k . This ~ expectation is confirmed by the appearance of IZstrong band at 795 cm-' in the infrared spectrum of saturated amine 5 and the absence of bands at 820 and 1022 cm-l, all of which identifies compound 5 as endo-2-N-dimethylaminobicyclo[2.2.1]he~tane.~ Since the product obtained by reduction of the iminium perchlorate salt of 3 with excess hexamethylenimine is identical with that obtained by reduction with lithium aluminum hydride, 1 the product must be endo-2-N-hexamethyleniminobicyclo [2.2.1]heptane ( 6 ) . It would appear, therefore, that a hydrogen is added to thle less sterically hindered ezo side of a bicyclic enamine during reduction by a secondary amine also. The reaction of N-Zbicyclo [2.2.1lheptylidene-2'methylhexamethy1t:niminium perchlorate (7) with ex(3) C. Kaiser, A. Burger, L. Zirngibl, C. S. Davis, and C. L. Zirkle, J. OW. Chem., 27, 768 (1962). (4) S. Beckmann and R,. Mezger, Chem. Ber., 89, 2738 (1956). (5) -4. C. Cope, E. Ciganek, and N.A. LeBe1, J. Am. Chem. Sac., 81, 2799 (1959).
10
11
12
pound was characterized by acid hydrolysis to aniline and benzaldehyde followed by identification of the aniline. Thus a secondary amine is oxidized to an imine by an enamine in this general type of reaction. From these data a mechanism can be written for the reaction using the iminium salt of norcamphor and hexamethylenimine (13) as an example. A nucleophilic attack of a hydride ion from the a carbon
13
I
14
(6) F. F. Blioke and N. J. Doorsnbos, ibid., 76, 2317 (1954). (7) J. H.Boyer and F. C. Canter, ibid., 77, 3287 (1955).
NOTES
FEBRUARY 1967
475
TABLE I YIELDSAND PHYSICAL PROPERTIES OF SATURATED TERTIARY AMINESFROM IMINIUM SALTS Component ketone to form iminium perchlorate
Sec- Reflux ondary time, aminea hr
Yield,
%
n tD
O C
%-
--C,
t,
Bp, " C (mm)
Formula
Calcd
%-
--H, Found
Calcd
Found
... ... ... Norcamphor 1 63 60b .. ... ... ... ,.. Norcamphor 2 15 25 84-85 ( 0 . 5 ) 1.4838 32 CuHzeClN04' 54.62 54.83 8,51 8,49 3 46 48 9 5 ( 0 . 5 ) 1.4681 Norcamphor 26 CisHzsN 80.64 80.91 13.08 13.23 4 65 62 131(0.3) 1.5563 Norcamphor 28 CieHziN 84.52 84.36 9.31 9.35 ... ... ... Cyclohexanone 5 18 28 103-105(15) 1.4798 26 d ... ... ... Cyclohexanone 6 40 0 , . ... ... ... ... ... ... ... Cyclohexanone 7 18 49 193(0.7) 1.6216 25 ... ... Cyclopentanone 8 88 19 108.5-109(15) ., , .. e ... ... ... ... ... ... Cyclopentanone 1 20 27* .. ... ... ... ... a The same secondary amine is used in the iminium salt and as the reducing agent. 1 = hexamethylenimine, 2 = 2-methylhexamethylenimine, 3 = di-n-butylamine, 4 = tetrahydroisoquinoline, 5 = pyrrolidine, 6 = morpholine, 7 = N-benzylamine, 8 = piperidine. b The physical properties of these amines are described in ref 1. e Identified as its perchlorate salt, mp 267-267" dec. d Reported previously by J. W . B. Reesor and G. F. Wright [J.Org. Chem., 22,375 (1957)l as bp 209-210, @D 1.4823. Identified by its picrate, mp 165-166" (lit. 166-166.5'). e Reported previously by J. I. Jones, J . Chem. Soc., 1392 (1950). Identified by its picrate, mp 171.5172" (lit. mp 174"). I
.
.
17
I6
18
19
.
I
of the secondary amine on the a carbon of the iminium salt results in the production of a saturated tertiary amine and carboniuni ion (14). Carbonium ion 14 then releases a proton t o produce imine 15. The reduction of enamines is not limited to the cyclic hexamethylenimine, but the cyclic amines, pyrrolidine, piperidine, and tetrahydroisoquinoline (17) , are also capable of reducing action. 3Iorpholine does not exhibit this reducing property. The aromatic and aliphatic amines represented by N-benaylaniline and din-butylamine, respectively, likewise show the ability to reduce enamines. The yields of these reactions along with the physical properties of the saturated amine products are surnmariaed in Table I. A tertiary amine, S-methylhexamethyleniminel8 does not show the reducing property toward N2-bicyclo [2.2.1 ]heptylidenehexamethyleniminium perchlorate (13) that the corresponding secondary amine, hexamethylenimine, does. Thus apparently tertiary amines do not undergo hydride transfer reactions with enamines as they do with the triphenylmethyl carbonium ion. 3,4-Dihydroisoquinoline (19) is not the oxidation product formed when iminium salt 16 is treated with
clod-
.
of the ultraviolet and infrared spectra of this product with that reported for isoquinoline1° and 3,4-dihydroisoquinoline. It was demonstrated that 3,4-dihydroisoquinoline itself would not reduce enamines by treating iminium salt 16 with 3,4-dihydroisoquinoline (19) and finding that no saturated amine 18 was produced. Apparently the isoquinoline is produced in the first reaction by a disproportionation of the 3,4-dihydroisoquinoline produced initially into tetrahydroisoquinoline (17) and isoquinoline (20). This type of disproportionation reaction with 3,4-dihydroisoquinolines has been observed bef0re.l' Experimental Section The spectra reported in this work were recorded on a PerkinElmer Model 137 infrared spectrometer and a Beckman DK-2 spectrometer. The analyses were carried out by Schwarzkopf Microanalytical Laboratory, Woodside, N . Y. All of the iminium perchlorate salts were prepared by the method of Leonard and Paukstelis.12 2-Methylhexamethylenimine was prepared by the method of Blicke and Doorenbos.B N-Methylhexamethylenimine8 was prepared by treating hexamethylenimine with formic acid and formaldehyde. AI-2Methylazacycloheptene was produced by the method of Boyer and Canter 3,4-Dihydroisoquinoline was made by treating the formate of p-phenethylamine with polyphosphoric acid as described by Pratt, Rice, and Luckenbaugh.13 General Procedure for Reduction of Enamines with Secondary Amines.-A stirred mixture of an excess of secondary amine and iminium salt was refluxed for a period of time. An excess of 6 N hydrochloric acid was added, and the reaction mixture was refluxed for 2 hr. The solution was made strongly basic with 6 -V sodium hydroxide and extracted several times with diethyl ether, The combined ether extracts were dried over anhydrous magnesium sulfate and filtered, solvent was removed, and the residual oil was distilled through a fractionating column. Nonacid-Catalyzed Reactions.-No saturated amine product is obtained by treatment of 1-N-hexamethyleniminecyclopentene~4 or 2-N-hexamethyleniminobicyclo [2.2.1]hept-2-enel with excess hexamethylenimine in the manner described above when no acid catalyst is present.
.'
20
excess tetrahydroisoquinoline (17), but rather isoquinoline (20) is formed. This was shown by comparison (8) R . Lukes and J. Malek, Collection Czech. Chem. Commun., 16, 23 (1951). (9) R . Damico and C. D. Elroaddus, J . Org. Chem., 31, 1607 (1966).
(10) H. Freiser and W. L . Glowacki, J . Am. Chem. Soc., 7 1 , 514 (1949). (11) C. I. Brodrick and W. F. Short, J . Chem. Soc., 2587 (1949). (12) N . J. Leonard and J. V. Paukstelis, J . Org. Chem., 18, 3021 (1963). (13) E. F. Pratt, R . G. Rice, and R . W. Luckenbaugh, J. A m . Chem. Soc., 79, 1212 (1957). (14) M. A . Volodina, V. G . Mishina, A . P. Terent'ev, and G. V. Kiryushkina, Zh. Obshch. K h i m . , 31, 1922 (1962).
